Our application UX is potentially sensitive to the uncle/ommer rate. I have been trying to understand the impact of PoS change on this phenomenon, but have not found any recent information that appears definitive, and old discussions are conflicting:
- One simple reading is that uncles should not exist (What are the anticipated economics of moving from PoW to PoS?)
- Yet, something akin to uncles does come up (Are there still uncles with proof-of-stake?)
What is the expected mechanics of uncle/ommer block creation and rate following the merge?
I read the referenced comment on the block content.
It is indeed clear that uncles will no longer exist as block entries. Still, it does appear that, from the user perspective, a substantially similar phenomenon can occur if a validator fails to propose within their slot or other failure modes cause disagreements among parts of the network and subsequent reorgs.
In any case, the following spells out how uncles arise under PoS, and why it should be orders of magnitude rarer: https://www.reddit.com/r/ethdev/comments/vgzeae/what_is_the_uncleommer_block_mechanics_under_pos/